Journal Article

FZJ-2026-05915

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png VIPR: Versatile inverse problem software framework

Rustamov, J. ; Creutzburg, S. ; Bornschein, J. ; Bussmann, M. ; Ganeva, M.FZJ* ; Gerlach, A. ; Häusler, S.FZJ* ; Helm, B. ; Hinderhofer, A. ; Juzak, R. ; Koutsioumpas, A.FZJ* ; Löser, R. ; Munteanu, V. ; Pandit, V. ; Schramm, U. ; Schreiber, F. ; Mothes, N. ; Kelling, J. (Corresponding author)

2026
Elsevier Amsterdam [u.a.]

This record in other databases:  

Please use a persistent id in citations: doi:10.1016/j.softx.2026.103163  doi:10.34734/FZJ-2026-05915

Abstract: VIPR (Versatile Inverse Problem Software Framework) is a plugin-based software framework for building reproducible machine-learning workflows for scientific inverse problems. These problems arise when researchers need to reconstruct physical properties of a sample from incomplete or indirect measurements, a common challenge in X-ray and neutron scattering experiments where only intensity measurements are available and phase information is lost. VIPR addresses such ill-posed reconstruction tasks where direct inversion is often impossible. VIPR implements a modular microkernel architecture with domain-specific plugins to produce configurable machine learning workflows, including both deterministic and probabilistic models. Workflows are defined via declarative YAML configurations and can be executed through a command-line interface or a containerized web application. For a given experimental dataset, VIPR produces standardized analysis artifacts, including visualizations and statistical summaries. This article demonstrates VIPR for X-ray and neutron reflectometry and for reconstructing PEMFC operating parameters from sensor data. Both applications use domain-specific plugins with the same core inference workflow.

Keyword(s): Instrument and Method Development (1st) ; Information and Communication (1st) ; Condensed Matter Physics (2nd) ; Instrument and Method Development (2nd)

---

Contributing Institute(s):

1. JCNS-FRM-II (JCNS-FRM-II)
2. JCNS-4 (JCNS-4)
3. Heinz Maier-Leibnitz Zentrum (MLZ)

Research Program(s):

1. 632 - Materials – Quantum, Complex and Functional Materials (POF4-632) (POF4-632)
2. 6G4 - Jülich Centre for Neutron Research (JCNS) (FZJ) (POF4-6G4) (POF4-6G4)

Experiment(s):

1. MARIA: Magnetic reflectometer with high incident angle (NL5N)

Appears in the scientific report 2026

---

Database coverage:
; ; ; ; Article Processing Charges ; Clarivate Analytics Master Journal List ; Current Contents - Engineering, Computing and Technology ; DOAJ Seal ; Essential Science Indicators ; Fees ; IF < 5 ; JCR ; SCOPUS ; Science Citation Index Expanded ; Web of Science Core Collection

---